Use of Medical Care during Pandemic (H1N1) 2009, Navarre, Spain

Of 233 households with laboratory-confirmed pandemic (H1N1) 2009 in Navarre, Spain, only 64% (107/166) of contacts with influenza-like illness had sought medical care. This value was lower for adults (53%, 39/74) than for children <15 years of age (74%, 68/92), as well as for those with cases secondary to another household case (58%, 64/111).

population, and most persons have various episodes of this disease during their lifetime (1,2). Infl uenza is typically characterized by respiratory symptoms and general symptoms such as fever, malaise, and myalgia (2,3). Most cases resolve without complications within 1 week (4). The epidemiologic context facilitates diagnosis of infl uenza because it usually occurs in seasonal waves (5). For these reasons, some persons with infl uenza do not seek medical care.
This proportion of infl uenza cases is hidden from the health system and epidemiologic surveillance, leading to underestimation of the true extent of the affected population. The purpose of this study was to estimate the proportion of persons with infl uenza who sought medical care in Navarre, Spain, during the 2009-10 infl uenza season and identify characteristics that differentiate persons who sought or did not seek medical care.

The Study
The study protocol was reviewed and approved by the Ethical Committee of the Barcelona Health Institute. During the 2009-10 infl uenza season, the Sentinel Network of Primary Care Physicians and Pediatricians of Navarre obtained nasopharyngeal swab specimens from patients with infl uenza-like illness (ILI) for virologic confi rmation of infl uenza by using standard real-time reverse transcription PCR or cell culture (6).
In early 2010, a trained nurse telephoned households of persons who had had laboratory-confi rmed pandemic (H1N1) 2009 in October and November 2009 while pandemic infl uenza was active in the area (index cases). When no response was received, the calls were repeated 5 times on different days and at different times. The interview was conducted by using a structured questionnaire. For each household, an attempt was made to talk to the adult who was primarily responsible for the health issues of those who lived there, usually the mother or father. When possible, other adults in the household were also interviewed.
Detailed information was obtained about index casepatients and all other persons living in the same household (contacts) with regard to sociodemographic data, medical history, infl uenza symptoms, and whether he or she had sought medical care. A question was asked about each of the following signs and symptoms: fever, cough, sore throat, headache, muscular or joint pain, nasal congestion, and vomiting.
This analysis included all household residents who had ILI for <7 days with respect to the onset of ILI in the index case-patient. ILI was defi ned as fever and either cough or sore throat in the absence of other diagnoses. The fi rst person with ILI in the household was considered the primary casepatient, and others, if any, were considered secondary casepatients. Persons with index cases who led us to contact the household were excluded from analysis because, given the study design, they had all sought medical care.
Of 252 households that met the inclusion criteria, 233 (92%) were successfully contacted, and all persons agreed to participate in the study. Of 668 household contacts of index cases, 188 (28%) persons had >1 infl uenza symptoms and 166 (25%) met the criteria for having ILI. The proportion of symptomatic cases that met the criteria for ILI was higher among children (94%, 92/98) than among adolescents and adults (82%, 74/90; p = 0.021).
No differences were detected by sex, rural/urban residence, country of origin, vaccination against seasonal infl uenza, smoking status, or presence of major chronic conditions. The percentage of secondary case-patients who sought medical care was the same regardless of whether the primary case-patient received antiviral treatment. Persons who had sought medical care had a mean ± SD of 4.0 ± 0.8 symptoms, which was similar to those who had not sought medical care (4.0 ± 0.7 symptoms; p = 0.688). None of the contacts had received antiviral prophylaxis or vaccine for pandemic (H1N1) 2009.
The frequency with which persons with ILI sought medical care, by age group and symptoms, is shown in Table 2. Frequency of symptoms did not differ between persons who had sought medical care and those who had not sought medical care.
Logistic regression analysis showed that seeking medical care was more frequent among children than among adolescents and adults (adjusted odds ratio 2.2, 95% CI 1.1-4.3; p = 0.019) and among primary case-patients than among secondary case-patients (adjusted odds ratio 2.2, 95% CI 1.0-4.8; p = 0.038). A similar proportion of contacts with any symptom but who did not meet ILI criteria sought medical care (59%, 13/22) as did those who met ILI criteria (64%, 107/166; p = 0.642).

Conclusions
We studied persons who had ILI and were household contacts of persons with laboratory-confi rmed infl uenza. It is likely that the cause of symptoms among persons with ILI was also infection with infl uenza virus. Approximately two thirds of these persons sought medical care. Care seeking was less frequent among adults and when there had already been another case within the household.
Epidemiologic surveillance systems and studies of infl uenza are typically based on cases of medically diagnosed ILI. Therefore, the proportion of persons not seeking medical care is usually unknown and not taken into account. Ideally, epidemiologic surveillance should consider these persons to avoid underestimating the actual magnitude of the disease (7). These persons may increase spread of the disease (8) and may also be more likely to selfmedicate, resulting in possible risks to their health (9,10). Persons who do not seek medical care do not contribute to the cost of infl uenza from the health system perspective, but they do from the point of view of society because they can result in lost work hours, more medications used, and increased need for care (11).  When infl uenza symptoms were present, adults and adolescents sought medical care less often than children. This fi nding may explain in part why the incidence of medically diagnosed ILI is usually much higher among children (4,6,12).
This study was conducted during the pandemic infl uenza (H1N1) 2009 season. Therefore, its results may not be generalizable to other infl uenza seasons (4). Nevertheless, the study was conducted when pandemic (H1N1) 2009 virus had already been circulating in the population for several months, the initial alarm had abated, and the level of medical care had returned to levels similar to that for seasonal infl uenza.
We included symptoms of infl uenza that were reported by concerned persons or concerned parents. Thus, persons with mild symptoms may be underrepresented in our analysis. The probability of making telephone contact may have been higher in households with more members. However, only 8% (19/252) of households were not contacted. Families in whom none of the members had sought medical care were not included in the study, which may overestimate the proportion of persons who sought medical care.
Persons with infl uenza who did not seek medical care should be taken into account in estimations of the actual incidence of infl uenza and its effect on the general population. These persons may have a major effect on transmission and should be considered in planning prevention and control measures and in evaluations of the effects of this disease. This study was supported by the Instituto de Salud Carlos III (GR09/0028 and GR09/0029) and by the I-MOVE Program supported by the European Centre for Disease Prevention and Control.
Dr Burgui is an epidemiologist at the Instituto de Salud Pública de Navarra, Pamplona, Spain. Her research interests are the epidemiology of infectious diseases and vaccines.